In January 2026, an economics researcher at RV University in Bengaluru, India, opened a journal rejection letter and found something unusual in the reviewer comments: her unpublished manuscript was “highly similar” to a paper that had already been published elsewhere. The paper it resembled was hers — stolen, resold to strangers as authorship slots on Telegram, and published eight months earlier under other people’s names. The case, first reported by Retraction Watch and followed through to retraction in a March 2026 update, is one of the more fully-documented examples of a specific paper-mill tactic: not fabricating research, but stealing a real, unpublished one and selling the byline.
What happened
The researcher, Vijayalakshmi S of RV University, had submitted the study to Discover Energy (Springer Nature) in January 2024; it was rejected after peer review. She had also presented the work at IIT Bombay’s Young Scholar Initiative conference. At some point between the rejection and her later discovery of the theft, the manuscript — or the data and framing underlying it — was taken and routed into a paper mill’s inventory.
In April 2024, a version of the stolen study was submitted to the Journal of Infrastructure, Policy and Development (JIPD), published by EnPress Publisher. It appeared in print in September 2024, listing a first author from Dhofar University in Oman and additional co-authors based in Oman, India, and Saudi Arabia — none of them connected to the original research.
The theft only came to light because Vijayalakshmi S posted about the plagiarism accusation against her own work on social media. Ashutosh Tiwari, who runs a counseling practice unrelated to academic publishing, searched Telegram using distinctive keywords from her paper and found the advertisement that had sold it. According to Retraction Watch’s reporting, the listing priced authorship as follows:
- First author: 15,000 Indian rupees (roughly $165)
- Second author: 10,000 INR (roughly $110)
- Remaining slots: 5,000 INR (roughly $55)
JIPD’s editorial team opened what it described as a “thorough investigation” and, by March 2026, removed the paper from its website entirely. No formal retraction notice was issued and no reason was posted — a departure from the Committee on Publication Ethics (COPE) retraction guidelines, which call for a public notice stating what was retracted and why. Separately, Scopus had already stopped indexing JIPD in January 2025, citing declining publication standards, months before this specific case surfaced.
Why this case is different from the usual paper-mill story
Most paper-mill coverage — including CASRAI’s own reporting on the BuyTheBy dataset — documents mills selling authorship on manuscripts they generated or templated themselves: tortured-phrase paraphrasing, formulaic data-free “research,” image reuse across supposedly unrelated papers. This case is a variant with a different, arguably more damaging supply chain: the underlying manuscript was real, competent, unpublished research, and the mill’s product was simply unauthorized access to it.
That distinction matters for how research administrators and integrity officers should think about exposure:
- Rejected and pre-print manuscripts are inventory, not dead ends. A paper rejected after peer review, or presented at a conference without a corresponding publication yet, sat in circulation long enough to be taken. Editors and reviewers routinely have access to unpublished manuscripts; conference presentations put full methodology and results in front of an audience before print. Either point is a plausible leak vector, and neither is currently treated as sensitive by most institutional workflows.
- The victim absorbs the reputational cost of the crime. Vijayalakshmi S’s own paper drew a plagiarism flag against her, because by the time she resubmitted, a version of her own work already existed in the literature under someone else’s name. This is the inversion that makes theft-based mills structurally worse than fabrication-based ones for the individual researcher: the honest author is the one who has to prove originality against a copy of her own work.
- Telegram functions as an open storefront, not a hidden one. The advertisement was found by an outsider to the field using ordinary keyword search on a public-facing Telegram channel — not through insider access or specialist tooling. This is consistent with what the BuyTheBy dataset documented at scale: paper-mill sales channels are often searchable and semi-public, and detection is frequently a matter of someone happening to look, not a systemic screening failure being closed.
- Enforcement remained slow and incomplete even after public exposure. JIPD pulled the paper without issuing a COPE-compliant retraction notice, and Scopus had already de-indexed the journal a year earlier for unrelated quality concerns. Two different accountability mechanisms (index delisting, retraction) each moved on their own schedule, neither of them fast, and neither producing a public record a future researcher or hiring committee could easily check.
What this means for institutions and journals
The case reinforces a point made across CASRAI’s existing paper-mill coverage: authorship-for-sale is not primarily a fabrication-detection problem, it is a provenance and access-control problem. Screening tools built to catch tortured phrases or duplicate images — the kind aggregated through the STM Integrity Hub — are of limited use against a manuscript that is, on its technical merits, genuine, unmanipulated research. The signal here was not textual; it was that the “new” authors had no verifiable connection to the underlying study.
Practical implications worth building into institutional and editorial practice:
- Authorship justification, not just authorship declaration. Requiring a short account of each listed author’s actual contribution (the discipline CASRAI’s own CRediT taxonomy is built around) creates a paper trail that is much harder to fabricate convincingly than a bare author list, and gives editors a concrete basis to challenge an implausible-looking author set.
- ORCID and institutional email verification at submission. None of the fraudulent co-authors in this case needed to demonstrate any prior connection to the manuscript’s subject matter or to each other. Requiring verified ORCID iDs and institutional (not personal-domain) contact emails at submission raises the cost of assembling a fabricated author list, without adding meaningful friction for legitimate co-authors.
- Treat rejected/withdrawn manuscripts as an access-control concern, not just a workflow dead end. Editorial systems, reviewer pools, and conference proceedings platforms that hold full unpublished manuscripts are a leak surface. This isn’t a reason to stop peer review or conference presentation, but it is a reason institutions should have some means of tracking a manuscript’s known circulation history if plagiarism is later alleged against the true author.
- Publish real, dated retraction notices. COPE’s retraction guidelines exist precisely so a case like this leaves a clear, searchable public record. A paper quietly removed from a journal’s website, with no notice, does not fully close the loop — it leaves the stolen version’s citation history, and any downstream papers that cited it, unresolved.
For the broader landscape this case sits inside — sanctioned paper-mill pricing data, publisher-side detection infrastructure, and the sector-wide coordination effort against the practice — see CASRAI’s coverage of the BuyTheBy authorship-pricing dataset, United2Act’s publisher coalition, the COPE Paper Mills Working Group, and the guide to paper-mill red flags in manuscript text.
Sources
This article draws on Retraction Watch’s original reporting (Dalmeet Singh Chawla, “Study is stolen, sold, published. Now the victim is accused of plagiarism,” January 26, 2026) and its follow-up (“Stolen economics study retracted following Retraction Watch coverage,” March 9, 2026). No details beyond what these two reports document are asserted here; where a detail (such as the exact retraction date or the full list of fraudulent co-authors) was not specified in the reporting, this article does not supply one.
